Atmospheric temperature profiling microwave radiometers heretofore known and/or utilized have employed a plurality of discrete Gunn diode oscillators for the local oscillator. In such cases the number of frequencies of which the radiometer receiver is capable, and the frequencies themselves, are fixed.
Moreover, Gunn oscillators have several characteristics related to frequency stability that are undesirable for use in observations between 50 and 60 GHz. Gunn oscillators have a poor frequency stability with physical temperature changes, up to two MHz/.degree.C. If Gunn oscillators are allowed to be at ambient temperature, this equates to as much as a 140 MHz shift over a -20.degree. C. to +50.degree. C. operating range. This results in less than desirable temperature profiling. Therefore, current uses of Gunn oscillators in such systems require careful temperature stabilization of the oscillator housing. As a consequence, these instruments are complex, unwieldy, and power consumptive. Additionally, warm-up times are significantly long. Even when temperature stabilized, Gunn oscillators will wander over many megahertz in frequency. This wandering has a grave effect upon the accuracy of temperature profiling. Further improvement could thus be utilized.
Now known methods of atmospheric water vapor profiling are dramatically effected by various weather conditions, some methods being ineffective in the presence of clouds. Many current techniques for profiling water vapor are not passive (for example, radiosonde and/or laser techniques) thereby introducing complexity and added expense to such measurements. New approached to such profiling could thus be utilized.
Radiometers are calibrated by establishing the gain and offset of the system. Offsets can be quite high, more than 700K, depending upon the receiver dark noise. The sky observables in any waveband have a range of less than 100K. Therefore, a small percentage change in receiver noise can induce a significant error in observables. Receiver stability, or frequent gain/offset evaluations of the receiver, are therefore necessary.
Most radiometer systems incorporate internal references in the form of hot and cold loads, targets of known temperature, and/or known noise sources. These references are used as absolute or as transfer calibration (near-term calibration) standards. Hot and cold loads are an incomplete calibration in that they do not include front end components of the radiometer such as the antenna system, dielectric window, and antenna isolators. Also, hot loads require an extrapolation of calibration data since hot loads are above ambient temperature and sky observables are at ambient temperature or below. It would be beneficial for calibration temperatures to span the range of observables. An improved calibration system for these types of systems could thus be utilized.
While under proper conditions and in selected wave-bands, data from "tipping curves" are quite resolute, absolute calibration of radiometers in the 20 to 35 GHz range can only be accomplished with "tipping curves" when the atmosphere is transparent enough (opacity is low enough) such that there is a significant change in sky brightness with zenith angle, and provided that the sky can be assumed horizontally stratified and uniform. If not, it has been found that up to 15% variations in brightness in these wave-bands will occur, resulting in erroneous calibration values. Further improvement in calibration is therefore required.